From Dr. Judith Curry’s Climate Etc.
Posted on September 25, 2019 by niclewis | 24 Comments
By Nic Lewis
The editors of Nature have retracted the Resplandy et al. paper.
Readers may recall that last autumn I wrote several article critiquing the Resplandy et al. (2018) ocean heat uptake study in Nature, which was based on measured changes in the O2/N2 ratio (δO2/N2) and CO2 atmospheric concentration. These were combined to produce an estimate (ΔAPOObs) of changes in atmospheric potential oxygen since 1991, from which they isolated a component (ΔAPOClimate) that can be used to estimate the change in ocean heat content. In four articles, here and here, here, and here, I set out why I thought the trend in ΔAPOClimate – and hence their ocean heat uptake estimate – was overstated, and its uncertainty greatly understated, essentially because of errors in their statistical methodology. The bulk of my criticisms were largely accepted by the authors of the study. However, it was evident from their related Realclimate article that in their submitted correction they had also made a change in an unconnected assumption, with the effect of offsetting much of the reduction in their ocean heat uptake estimate that correcting their statistical errors would have caused.
Nearly ten months have passed since then, without Nature publishing the authors’ correction.
However, Ruth Dixon has just spotted that the Resplandy et al. paper has today been retracted, at Nature’s request. This article at Retraction Watch covers the story. The Retraction Notice by the authors at Nature reads:
Shortly after publication, arising from comments from Nicholas Lewis, we realized that our reported uncertainties were underestimated owing to our treatment of certain systematic errors as random errors. In addition, we became aware of several smaller issues in our analysis of uncertainty. Although correcting these issues did not substantially change the central estimate of ocean warming, it led to a roughly fourfold increase in uncertainties, significantly weakening implications for an upward revision of ocean warming and climate sensitivity. Because of these weaker implications, the Nature editors asked for a Retraction, which we accept. Despite the revised uncertainties, our method remains valid and provides an estimate of ocean warming that is independent of the ocean data underpinning other approaches. The revised paper, with corrected uncertainties, will be submitted to another journal. The Retraction will contain a link to the new publication, if and when it is published.
I believe that this saga, as well as showing how ineffective journal peer review tends to be in spotting problematic issues in papers, illustrates the need for a much closer involvement of statisticians in climate science research. That was a point also made in one of the articles highlighted in Judith’s latest Week in Review post: Climate science needs professional statisticians [link].
Nicholas Lewis 25 September 2019
Nic Lewis,
Your report can’t be right because Michael Mann and his claque tell me the science is settled and everybody knows that only evil deniers dispute that / sarc off.
Richard
Water vapor deserves way more GHG respect.
Water vapor is the most important greenhouse gas.
http://web.gps.caltech.edu/~xun/course/GEOL1350/Lecture6.pdf
Give them some time. They’re still trying to figure out what that big yellow ball of gas is in the sky.
They’ll get around to less important things like the most important greenhouse gas in about 11 years, 11 months or so…
It’s not gas it’s plasma. As is 99% of the matter in the universe ( not counting the “dark” matter” of course because we can’t see it, can’t detect it and can’t measure anything interacting with it. But we know it’s their otherwise our equation would be wrong).
Is water vapor a “climate pollutant” then?
Don’t give them any ideas. I can just see someone getting a government grant to suck clouds of the sky to sequester them underground…
BTW, I always ask True Believers if we should do everything we can to lower the emissions of the greatest greenhouse gas. When they (of course) say yes, I ask them how much they think it would cost to dam up Niagara Falls…
Yes.
Personally I would like someone to ask the EPA to rule it pollution same as CO2 for same reasons. Why? Because I don’t think they will get away with calling it hydrogen pollution. And, I think it will make people start to question more.
If they declared hydrogen a pollution wouldn’t they have to declare oxygen also?
That will have to wait for the next little moppet to declare on the next Climate Strike…
I think its good that they don’t actually teach science to kids these days. Think what they’d do when they discover the Periodic Table. I mean, that’s just a list of poisons, that is…
“If they declared hydrogen a pollution wouldn’t they have to declare oxygen also?”
Not necessarily. N2 is not a pollutant. O2 is not a pollutant. NOx and O3 are pollutants.
” illustrates the need for a much closer involvement of statisticians in climate science research.” That so true and this also true, ” illustrates the need for a much closer involvement of statisticians in science research. Far to many research paper are trash due to the lack of good statisticians have a look at published paper long before they are published!
As I recall, that is precisely what Wegman said way back in 2006 about the “hockey stick”. Thirteen years late and nothing has changed. A lot of so-called settled “climate science” is still based on incorrect use of statistics.
And IT personnel. These quack are not coders, they don’t know a thing about models, schemas, entity relationships, data integrity, code integrity, testing, and application logic. Nothing.
I would guess that no more than 4-8 hrs would be needed to skewer and reveal as fraudulent, any of their models, apps, or code based programmes.
But of course ‘open science’ is not allowed. For we the peasants, including IT experts, are too stupid to understand the ‘settled science’ based on fraudulent apps and models.
Agreed. I work in the health data field as a statistician. Between us, the IT crew and the actual health experts, there is often much metaphorical blood spilled during what could only tangentially be called “discussions” now.
Unfortunately, as it goes right now, we are led by someone who really, REALLY likes infographics. So…everything looks purty but those of us who have serious doubts about the actual data are given a few minutes to vent, then put on “ignore”.
Oh, and I have to use a lovely database in Access that, literally, hasn’t been improved in 20 years.
But, damn, those blog posts look GREAT!.
Years ago I got enthused about statistics (I was a pathologist). I started reading medical articles carefully. Their statistical analyses were pathetic.
The problem is that the world is a messy place and it is hard to make predictions, esp about the future. Much of health research is based on observations, not carefully controlled laboratory experiments. There are any number of variables involved, many of which are not understood or taken into account. All most docs understand is a P value less that 0.05 proves something. So, they use statistics to make things seems more scientific. Fact is, almost nobody I have ever met understands, or even tries to understand, statistics. For example, I have never met another physician who understands that a P value less than 0.05 just means that there is a less than one chance in 20 that your results are not the results of random error, PROVIDING that you have rigorously shown that your result is a properly behaved random variable of a type whose behavior can be statistically described. I have yet to meet a physician who understand that you never prove a hypothesis, you simply fail to disprove it.
Etc.
I worked very closely for awhile with the big brains on a health care model. You can bore yourself asleep if you’d like to take a peek:
https://www.cihi.ca/sites/default/files/document/hig-infosheet-2018-en-web.pdf
Basically my job was to find old data (tree rings), explain why our old data was so crappy (tree rings don’t really tell you much), clean it up so that it was at least a bit presentable (proxies, proxies, and more proxies), find out how we could plausibly add it to our shiny new data (hockey stick the hell out of it) and then give it to the executive chefs (modelers) who would hocus-pocus it into something palatable we could spoon feed decision makers (IPCC) so that our PR folks could hype how important something was (MSM anointing Greta).
So, yeah, I think my insignificant little job in building this Frankenstein Monster gives me some insight into how the sausage is made…
(Apologies for the mixed metaphors…)
“Much of health research is based on observations, not carefully controlled laboratory experiments. There are any number of variables involved, many of which are not understood or taken into account. All most docs understand is a P value less that 0.05 proves something. ”
My PhD in immunology who studied statistics as an undergraduate will confirm this. Garbage going in, analyzed statistically, is still garbage when it comes out!
Indeed.
Funny how an old curmudgeonly engineer would be able to be so good at explaining this:
http://www.numberwatch.co.uk/Stuff.htm
(well, not funny, but predicable: he actually studied science when it was science).
Good Comment.
As a physician (and although not especially well-versed in statistics), I think that all studies should use nonparametric statistics; unless the data are “numerous/strong” enough to show that methods for analyzing random–Gaussian–distributions are applicable.
And none seem to understand that if you hunt for ps, and you have twenty or more ways of finding relationships, you probably just find a random one.
P hacking is disastrous for science and many p hacked papers are used to justify political policies.
As I recall, one of the Climategate emails was a cri de coeur from a database guy who despaired of ever making sense of their data.
For more amusement, consider the station metadata file that NOAA provides for the GHCN Monthly and GHCN Daily datasets. There is a column on each record named “GSN_FLAG”. When set, its value is…
“GSN”
Is that database design brilliance at its finest, or what?
This, in my experience with code written by PhD holders is true. I think the models are written in FORTRAN, which may be good or bad, depending on your view of the alternatives.
Consider the implications of MBH98 being withdrawn!
https://climateaudit.org/category/mbh98/
Hey, I know a semi-retired squash enthusiast they might be able to hire…
Mark Luhman-
Absolutely agree. Except I would say “…WAY, WAY, WAY, before they are published!” The statistician I worked with for many years, told me early in our association that I would get better results if I came to him while I was PLANNING a test series, not just when I was analyzing the data. I did from then on, and his advice kept me from wasting a lot of time and money, and lot of embarrassment.
True story: I once made a life-long enemy out of someone who constantly bragged about her Masters degrees and her attention to detail and never let you forget you were nowhere near as edumacated as she was (and going to be, she eventually got her PhD).
I noticed that a calculation she was using kept changing, which was expected, but way out of any parameter she SHOULD have looked into.
Turns out while she was constantly updating the numerator, the denominator was stuck.
We called her “F9” behind her back and she was of course promoted upwards…
Recalc?
I can imagine few things less palatable than chewing through 85% of your grant, only to have the statistician tell you at the data analysis stage that your methodology can’t find what you’re looking for.
“Climate science needs professional statisticians”
Wow – ain’t that the truth! Please tell MSM reporters too. Start with the establishment of historical mean temperatures.
M
That’s been glaringly obvious since M&M took down Mann and his co-conspirator’s hockey stick fraud over 15 years ago.
Exactly!
+1. And the MSM needs to hold horses when new papers get published.
Yep – the BBC gave this full headline prominence all over breakfast news and their webiste. I complained and pointed them to Nic Lewis. They just regurgitated the press release from Resplandy at the time.
Anyhow, great to have been able to go back to the BBC just now and complain again and point out the paper has been withdrawn and suggest that they need to show a prominent correction. They won’t, but there is some small satisfaction in being able to follow up the original complaint with an “I told you so” further complaint.
The BBC just doesn’t get the fact that peer review means almost nothing, and especially so in CliSci.
Old MSM business model: produce enough column inches (or TV minutes) so that the underwear ads don’t run all together. Put some thought into the content so that your senior writers actually had enough experience to be entitled to the word “senior”. Beat writers in particular would start at the bottom, learn the ropes and could truly be called experts after a decade or so.
Convergent MSM business model (i.e., AOL buying Time Warner): produce enough content so it can be scattered across all the corporate entities, which mean pulling those beat writers off their beat and pushing a camera in front of them. “Star pundits” were born. Start putting those pundits on TV, a half-dozen at the time where they can scream at each other before cutting to a commercial break then going out for dinner together.
New MS business model: hire 24 year old arts grads to re-write advocacy group press releases, scour social media for ideas, write a quick “hit”, press “send”, count the clicks and pray the measly ad revenue will enough to promote the “writer” to “senior writer”, although the only cost will be a new sign for the cubicle.
The BBC are bad but all the MSM are. They loudly trumpet “latest” research without any understanding that the newer the work, the more likely it us to be wrong . It’s the old stuff that has stood up to replication and use that is more likely to be right.
The BBC won’t even consider any complaint about a story that they published more than 30 days ago. Unless someone sues them, I presume.
Might explain how they couldn’t find a pedophile for decades…
@ Caligula Jones
When they tried, they fell off a Cliff…
They just need honest statisticians.
Statisticians normally have no idea of the physical sciences. If they are given garbage data in then no amount of them doing statistical analysis on it can result in anything except garbage out. Honesty in the statistician won’t help.
It will if the statistician is closely involved with the research, and not just a dispassionate observer.
“Although correcting these issues did not substantially change the central estimate of ocean warming, it led to a roughly fourfold increase in uncertainties”
Translates to:
“Correcting these issues quadrupled uncertainty and negates our assertions, but it did not substantially change our belief that we are right.“
That was my thoughts as well. The problem is, they don’t understand that the increased uncertainty does actually negate their assertions. They fundamentally don’t understand the difference between accuracy and precision. I have found a lot of educated people have a problem with the idea that if you don’t know something, you can’t say anything about it except to speculate (guess). They keep fooling themselves into thinking that they can overcome this obstacle by using fancy mathematics to estimate, extrapolate, smooth, filter, or average their way to a result they can report as knowledge. But in the end, it’s still just a guess – no better than throwing darts or reading goat entrails. So it’s no surprise that laypersons don’t spot these issues either; they have been “blinded by science”.
“… significantly weakening implications for an upward revision of ocean warming and climate sensitivity. Because of these weaker implications, the Nature editors asked for a Retraction …”
Huh? This reads as if Nature simply doesn’t want to publish non-alarming papers, no doubt a fact that would survive a fair Factcheck if such a thing existed. Sigh, a journal of integrity would insist that the paper is corrected and a new version published.
Still doesn’t stop the climate charlatans from making new pseudoscience claims about ocean warming:
Media:
Doesn’t matter. The propaganda-loving press for the Left has moved on to the next set of lies.
“Earth’s Oceans Are Getting Hotter And Higher, And It’s Accelerating”
NPR.org . 3:54, September 25, 20195:00 AM ET
https://www.npr.org/2019/09/25/755859707/earths-oceans-are-getting-hotter-and-higher-and-it-s-accelerating
The IPCC “Presser” garbage report for the ignorant who can’t reason for themselves:
First bald-faced lie:
– Half the warming observed occurred prior to 1950, when previous IPCC admits delta-CO2 was too low to attribute the observed warming then.
Second bald-faced lie:
– The oceans are becoming more productive, not less. Primary productivity is increasing. By every measure.
It is only in in silico models that ocean productivity decreased/decreases.
Third bald faced lie:
– “Causing sea level rise” is a lie. Sea level rise rate has remained steady since about 1880 at 2-3 mm/yr. Anthro-CO2/AGW hypothesis predicts it should be accelerating if oceans are warming at a faster rate than prior to 1950. Acceleration in SLR is not occurring, therefore neither acceleration of bulk ocean warming nor increasing glacial melt is occurring.
– Fourth bald-faced lie: Measured in cost per event, is a Yes: “extreme coastal events are becoming more severe.” But measured by an adjusted GDP and population growth in those areas, “cost per event” has fallen dramatically in the last 50 years along with the body count.
In the typical classic propaganda-style, the IPCC managed to get 4 bald-faced lies about CC/AGW in the last two sentences, whereas the first two sentences were arguably true.
The first rule of effective propaganda: Start out with unassailable “science” truths, then follow with the lies in subsequent sentences. Those who object to the utter BS of the statement are called out as “denying science.”
Both the NYTimes and the WaPo carried major science write-ups on the original Replandy, eta al. paper, only to “update” those stories on-line, after Nic Lewis pointed out the errors. ( I would add this is not Ms Resplandy’s first major data problem with a paper under her name as first author.)
But unless you search back 12 months for those old articles, you would not know that.
And neither WaPo nor the NYT will actually publish a new article for their today’s readers showing how their past readers were originally misled by bad science only to be retracted.
That is why journalism today is D-E-A-D.
“A lie can travel half-way around world, before the truth can get its boots on.”
– by someone not me.
Did you add the space to over whelming, or did they remove it?
“Global warming has already reached 1°C above the pre-industrial level, due to past and current greenhouse gas emissions.”
This is “the more than 100%” issue.
“There is overwhelming evidence that this is resulting in profound consequences for ecosystems and people.”
We’re not so much scared about the profound, but the harmful. There is, for example, profound greening going on, large enough to feed millions of people. There is a good chance cold-related deaths decrease.
“The ocean is warmer, more acidic and less productive.”
Less alkaline, very slightly warmer, and productivity is a thing you’d not capable of measuring independent of fishing and other human influence.
“Melting glaciers and ice sheets are causing sea level rise,”
So “are causing” at any moment now, trust me, not “have caused” anything of significance.
” and coastal extreme events are becoming more severe.”
I’m sure they have so good large-scale evidence on this. /sarc
Hugs,
I did a Copy-paste from the IPCC statement .pdf which lost all the spaces. I then added them back in. The “over whelming” extra space is my mistake. An artifact of a fallible human having to deal with technology.
That’s what evil doers do.
Alternatively there is the attitude of the BBC which refuses to make corrections or even consider issues arising for the benefit of future articles unless complaint is made within 30 days. The errors remain on the internet as “you can trust the BBC” gospel.
The BBC does not allow comments. They are not SO stupid.
A good thing to do if you don’t want to bother with the great unwashed finding fault with your life’s work, of course….
At the BBC is public, it should probably allow the public to have some input, but as I am no longer a British subject subjected to enforced fees…not my problem.
An interesting example of this is when the Toronto Star stopped allowing comments. You can understand, perhaps, from a business end of it. You’d have to always worry about libel, abuse, spamming, etc.
But…when I used to comment there, I would (to my eternal regret), “pop in” to check if anyone had read my comment, commented on my comment, etc. Maybe…dozens of times a day (you know…someone was wrong on the internet!!!!).
Now…not so often.
So, if you want clicks, open yer combox….
In their superior wisdom, they intend to educate the public, not to listen to the public. Pretty much like Greta. How do you dare to demand to be listened to?
★★★★★
+1,000
It’s the “WaPoo.”
Great! I have had to say for years that I am not a Scientist but I am a Statistician and thus am amply qualified to critique claims which i believe are largely opinions and not true Science.
Lack of error bars, correlation coefficients, homogenized data etc etc.
Good old Stats just might sink a few. Maybe we should ask Michael Mann about that?
Since energy emitted by CO2 can only penetrate a few microns I dont think the ocean is capable of ‘uptaking’ heat.
In fact all IR poorly penetrates water, IR from CO2 the worst, only visible does to any degree, so almost all ocean warming is due to visible radiation.
People will say ‘ah, SAGE Tangaroa’ but look at what they used. Cloudy days, 100 watts of downwelling IR, at a frequency different from CO2, and the response was a 0.02 C increase a few CM deeper.
That is tiny.
So while the 3 watts from CO2 might cause a warming of 0.0006 C, that warming is due to retained energy from visible radiation, not uptake of IR.
Like a dam of water. Adding more bricks to the lip raises the water level, but the increase in the dam is water, not bricks.
Ocean heat uptake is a myth. A physical impossibility.
As IR doesn’t penetrate deeply into water, one must ask, “Where then is the energy going?”
What is not being reflected (some is) must be absorbed. The fact that is doesn’t penetrate very deeply only means that its energy must primarily be warming the surface, which would have a large effect on surface evaporation and hence water vapor.
Whether extra IR is absorbed at the water surface, or whether IR is retarded from leaving the water surface, it ought to register a local measurable temperature change before that excess heat can ever progress into the deep oceans. The evidence is not there that I have seen.
That’s one good argument against Trenberth’s missing heat turning up in the abyss.
It isnt ‘going’ anywhere, it stays in the atmosphere. Why do you have to assume that just because it is there it is absorbed in some way?
Congratulations to Nic Lewis! Great! Real science wins.
Yes. Congratulations. Nic Lewis. It was not easy.
Respect to Nic Lewis for his dignified approach to correcting systematic errors in published papers.
One would think the authors of these papers would be smart enough to consult with a counterpart in the Math Department.
Although I think highly of Dr. Judith Curry and her work, I would not be in such a hurry to elevate statistics. Remember the Mark Twain comment (borrowed from Benjamin Desrali) “there are 3 types of lies: lies, damned lies, and statistics”. This comment was directed against using statistics to draw inferences not inherently in the data, wherein the data was not suitable for that inference. We need more scientists who present good data, and add an interpretation to it, and remain interactive and introspective with the reviews of it.
Agreed, statisticians tend to know a lot about statistics, and to be evangelical about its power (especially those strange Bayesian gurus), but can they discriminate between its undoubted power when applied to a series of coin tosses or measurements of the position of a satellite, and its feebleness when applied to climate data?
I would counter with a different expression: “Figures don’t lie, but Liars figure.” Properly used, statistics is all about drawing inferences that ARE inherently in the data, and discarding “inferences not inherently in the data”. It is not about torturing the numbers until they give you the answer you want. Proper unbiased analysis of the data is required if the scientific method is to work. The name for this unbiased analysis is statistics.
Nice try, Geoff, but statistics often fail to display reality. I give you an example: I buy ice cream in a large square plastic container, with a tab at one of the four corners that you break off to open the container. When I pick up the container, how often will I look first at the broken area to open it? I’m guessing you think 25% of the time, but here’s reality, I need to turn the damn thing four times to get to the tab. There! Reality!
Ron Long,
Climatology is the study of average weather (see e.g. the IPCC bibliography).
Averages are statistical constructs. Climate is a set of averages.
You may be right when you assert,
“We need more scientists who present good data, and add an interpretation to it, and remain interactive and introspective with the reviews of it.”
But that data only becomes part of – or relevant to – climatology when it is assessed, processed or compared using appropriate statistical procedures. Climate
You also say, “statistics often fail to display reality”. True, but so what?
The average height of an Englishman has no reality: my height is not the average of anything,. But, although the average height of an Englishmen has no reality, the average height and statistical distribution of heights of Englishmen are valuable information for clothes wholesalers wanting to sell to Englishmen.
Richard
The trouble is that scientists often do not understand the difference between “good data” and bad data.
A professional statistician cannot be of use unless he/she/it fully understands the subject and the nature of the problem to which the statistical tests are to be applied.
I was taught statistics from two sources. At university, as a beautiful extension of mathematics (random variable).
Later, in my professional apprenticeship, to derive and analyse any data from original sources which might be useful in furthering understanding, testing and, ultimately, forecasting. The first important step was the derivation of the data.
How many scientific tests are established without the advice of a professional statistician to point out the scope for errors in the collection of the data, spurious accuracy, internal correlation, noise, to name but a few?
But equally important was the preliminary analysis. It is nearly always at this stage that amateur statisticians, expert scientists in their own field, usually fail.
It is no surprise to me that as a professional statistician Nic Lewis, who has taken the trouble to learn the background of climate science can, and does, so often shoot down “peer-reviewed” seminal papers. Long may he continue to do it.
Nice Nick.
Actually not bad by Nature either. Even though the gears turn slowly.
Thanks, Jeff
Kudos to Nic Lewis. I read that the authors are standing by their flawed methodology, even with its reduced confidence levels (of course; what does quality matter to faux scientists?; it’s just an impediment to “progress”), and will re-publish in a lesser journal.
Good work Nic Lewis! I’m very thankful that there are people like you out there, that are willing to take on the pal review propaganda machine!
Despite the retraction, the authors are still insisting they were right…
https://retractionwatch.com/2019/09/25/nature-paper-on-ocean-warming-retracted/
When the error was corrected by Nic Lewis, the heat uptake dropped like a rock…
https://judithcurry.com/2018/11/06/a-major-problem-with-the-resplandy-et-al-ocean-heat-uptake-paper/
Yet Keeling says, “Despite the revised uncertainties, our method remains valid and provides an estimate of ocean warming that is independent of the ocean data underpinning other approaches.” Unbelievable.
Data is such a hindrance to progress.
“An estimate of ocean warming that is “independent” of the ocean DATA underpinning other approaches.”
In other words, when measurements of ocean temperature don’t supply the answer you want, invent another answer backed by assumptions and inferences, to “find” what they expect to find, given their incorrect hypotheses and models and imaginary “feedbacks” for which no empirical evidence exists. So once again, confirmation bias on steroids, just in time for another propaganda headline during the latest “climate” meetings.
AND they will republish, with the fact that the uncertainties render their “result” meaningless bullshit being ignored.
Unbelievable, indeed.
“So once again, confirmation bias on steroids, just in time for another propaganda headline during the latest “climate” meetings.”
Does anyone know if the just released IPCC report on the state of the oceans referenced claims made in this study?
The Paper is referenced in Chapter 5.
https://report.ipcc.ch/srocc/pdf/SROCC_FinalDraft_Chapter5.pdf
I noticed that as well today.
I my not so humble opinion, it could explain why the retraction was only done yesterday.
As they where awaiting the release of the IPCC report which references it before doing the retraction.
A geologist working down the hall from me used to always say “falling like a head shot duck” 😉
Much higher rate of descent than your standard rock.
I think we can consider Nature to be the New York Times of scientific publications. It was once great, but now only good for lining the cat litter box.
“correcting these issues did not substantially change the central estimate of ocean warming, it led to a roughly fourfold increase in uncertainties, significantly weakening implications for an upward revision of ocean warming and climate sensitivity”
Ya think?!?
Right after hiring a competent statistician (or anyone with a B from an eight-grade math course from the 1970s) they should hire a real comedy writer. This stuff isn’t funny in the way they think it is.
Well Done Nick.
The question would be “can current climate science survive an invasion of statisticians?” Given what I’ve read from M&M onward, I think not. However, a new, more useful climate science might arise from the ashes, and that would be truly useful. Worth a shot, anyway…
‘Despite being wrong, we’re still right.’ Sounds like they are graduates of the Dan Rather School of Communications.
I.e. the old saw: “the idea works in practice, but I don’t believe it because it doesn’t work in theory.”
As I’ve said, if more people were better at math, they’d be much less worried about most things the MSM hypes and perhaps much more worried about things the MSM ignores.
Something that “doesn’t affect our estimates” yet somehow quadruples your uncertainties” would get you fired in most other jobs, outside of economist or psychic.
I would suggest that Ph.D. programs need to rethink their statistics classes to one that focuses more on how to interpret professional statistical results. When I was in my masters program (I was doing research at that level), I took a Ph.D. level statistical methods class. I learned how to do an ANOVA with just a calculator. At that time I also bought Statview and used it during the data analysis phase of my research. If they are still teaching statistics that way to Ph.D. candidates, they should stop. It gives them the impression that they can do their own analysis. My research was very simple. The paper described here was certainly at the upper end of complexity and demanded for more advanced data methods. And I believe that points back to the impression given to Ph.D. candidates that they can do their own analysis.
Far too many PhD programs today don’t even require classes in statistics. “You can always find a math major to do the math!” And we expect peer reviewers to actually critique statistical analyses?
Nature seems to have become quite politicized. Just look at the September 19 issue that is dedicated to climate activism
I’m wondering how the folks that cited Resplandy et al. would revisit their own studies, now that a leg has been removed from their table of citations.
Also, I found it almost funny that the team found an ‘unconnected assumption’ that could be revised. Where was peer review in calling out the propriety of the assumption in the first place?
OH! The IPCC got caught in this too! Goody, goody.
So far I haven’t noticed anyone make a connection between the various explanations for “the pause”(or “ocean heat uptake” as one rationale for why a pause could occur), and that they invalidate the assumption of energy balance at TOA that is fundamental to climate models.
Personally I don’t see why the assumption has merit. As soon as solar energy enters the hypothetical sphere surrounding earth it can be transformed in a myriad of ways with no need for it to leave via the sphere via LWIR to maintain a stable atmospheric temperature.
To give just one example behind my thought process, a gigaton of ice situated 1000 meters above sea level in Greenland has more potential energy than a gigaton of ice that is bobbing about the ocean in icebergs. Those net amounts are likely to change day by day and year over year. Solar energy is the primary force that causes a change that begins with evaporation.
What happens within the sphere is more of an accounting (inventory) problem. Modelers can pretend to solve an inventory problem with parameterizations and sophisticated partial differential equations that are useful for solving physics problems, but until wrestling with the many ways energy can transfer and be stored (as well as originate in varying amounts from within), I can’t see how model results could be credible.
Congrats to Nic. This is what needs to be pursued in order to get the word out that models are telling us nothing.
Putting research thru a statistician might help in some things when it comes to analyzing data but there are other areas where climate scientists, and I suspect statisticians, too also have no clue.
The very data the scientists are dealing with have measurement errors, uncertainty, and basic reliability issues.
The very fact that averaging data is allowed to add a digit or two of precision would be unthinkable to experimental physicists and chemists and to most any college student in these areas. Claiming that averaging temperatures results in more accurate data when different locations have different standard deviations is turning error of the means on its head.
At least these authors mentioned uncertainty. However, I don’t think they really believe that this affects the reliability of their results. It’s too bad editors and reviewers don’t require an error budget be determined for each study using data that includes uncertainty, and measurement errors.
It looks like a lot of climate papers should have an expert in the real physical sciences that are familiar with taking measurements and assessing the propagation of errors be included as an author. I would hope that these scientists would not want papers with these kinds of errors to have their name included.
Now what’s the probability this will be reported in the NYT, LAT, BBC, or NPR?
Our analysis of ocean heat content
https://tambonthongchai.com/2018/10/06/ohc/
Nic,
Well done. You have a right to be pleased with your contribution.
I would also congratulate the authors for accepting the necessity of radical amendment. I can think of numerous examples where authors become entrenched in defending their views even when glaring problems are pointed out. I notice that the Marvel et al paper:-“Implications for climate sensitivity from the response to individual forcings” has clocked up over 50 citations. The Marotzke and Forster paper has clocked up over 100. When the peer review process fails, the mechanisms for stopping the propagation of bad science seem wholly inadequate. The Resplandy retraction is a rare but welcome success in this regard.
Was that the same Marvel of comic book fame?